Compare prompts

The Compare feature lets you see how a different prompt, model, or a parameter setting changes the model's output. You can view each of the prompts and their responses side by side to compare and analyze in the following ways:

• With a new prompt.
• With another saved prompt.
• With a ground truth.

Before you begin

To access the Compare feature, follow these steps:

1. In the Google Cloud console, go to the Create promptpage.

   Go to Create prompt

2. Select Compare. The Comparepage appears.

Create a prompt in the Compare feature

On the Comparepage, you can create a prompt before selecting another prompt to compare results.

To create a prompt, follow these steps:

1. In the New Promptfield, enter your prompt.

2. Click Submit prompts. The model's response appears below the prompt text that you entered.

3. Click Save as new. A Save promptdialog appears.

4. Enter the name of your new prompt in the Prompt namefield.

5. Select your region in the Regionfield, or leave it as the default region.

6. If a customer-managed encryption key (CMEK) applies, do the following:

   1. Select the Customer-managed encryption key (CMEK)checkbox.
   2. Select a key from the Select a Cloud KMS keyfield.

7. Click Save, which saves your prompt in the list of prompts to use on the Compare saved promptpage.

8. Click Submit promptsto compare the prompts and their responses.

You can update your prompts, and save updated versions as new prompts.

Compare with a new prompt

To compare your saved prompt with a new prompt, follow these steps:

1. Click Compare new prompt . A Compare pane appears.
2. Optional: Click Switch model to use a different model from the default model.
3. Optional: Expand Outputs .

Outputs:

1. Optional: If you want the model to output in a specific format such as JSON, click the Structured output toggle. After you select Structured output , the Grounding options are turned off, because grounding isn't supported with structured output.
2. Optional: Change the Thinking budget to one of the following options:
• Auto : The model only thinks when it needs to. The model adjusts how much it thinks or analyzes a situation based on what's needed at the time.
• Manual : You can adjust the thinking budget tokens.
• Off : No thinking or budgets are used.

4. Optional: Expand Tools .

Tools:

1. Select one of the following options:
• Grounding: Google : Grounding with Google Search or Google Maps.
• Grounding: Your data : Grounding with Vertex AI RAG Engine, Vertex AI Search or Elasticsearch.
1. If you select Grounding: Your data , select the data source that you want to use.

5. Optional: Expand Advanced :

Advanced:

1. Select Region .
2. Select Safety Filter Settings . A dialog appears. Keep the default of Off , or you can specify Block few , Block some , or Block most for each of the following options:
• Hate speech : Negative or harmful comments targeting identity or protected attributes.
• Dangerous content : Promotes or enables access to harmful goods, services, and activities.
• Sexually explicit content : Contains references to sexual acts or other lewd content.
• Harassment content : Malicious, intimidating, bullying, or abusive comments targeting another individual.
3. Click Save to save the settings and close the dialog.
4. Select the temperature from the Temperature field. The temperature controls the randomness in token selection. A lower temperature is good when you expect a true or correct response. A higher temperature can lead to diverse or unexpected results.
5. Select the output token limit from the Output token limit field. Output token limit determines the maximum amount of text output from one prompt. A token is approximately four characters.
6. Select the maximum responses from the Max responses field. If the maximum number of model responses generated per prompt. Because of safety filters or other policies, responses can still be blocked.
7. Select a value from the Top-P field. The Top-p changes how the model selects tokens for output.
8. Click toggle on the Stream model responses field. If selected, the responses are printed as they're generated.
9. Enter a stop sequence in the Add stop sequence field. Press Enter after each sequence.

6. Click Save to save changes to your settings.
7. Click Apply .
8. Click Submit prompts to compare the prompts and their responses.

For more information on token limits for each model, see Control the thinking budget .

Compare with another saved prompt

To compare your saved prompt with another saved prompt, follow these steps:

1. Click Compare saved prompt . The Existing Prompt pane appears.

2. Choose up to two existing prompts to compare.

   1. Select a Prompt name . If you have many prompts in your list, click in the Filter field, and select the property that you want to filter by. Enter a value, and press Enter .
   2. Click Apply . The Comparepage displays the prompt that you've selected alongside other prompts that you've created or selected for comparison.

3. Click Submit promptsto compare the prompts and their responses.

Compare with a ground truth

Ground truth is your preferred answer to the prompt. All other model responses are evaluated against the ground truth answer.

To compare your saved prompt with a ground truth, follow these steps:

1. Click Ground truth . The Ground truth pane appears.
2. Enter your ground truth to generate additional evaluation metrics.
3. Click Save to save the ground truth.
4. Click Submit promptsto compare the prompts and their responses.

The evaluation metrics that are generated when you compare a prompt with a ground truth aren't affected by the region that you select.

What's next

• Explore more examples of prompts in the Prompt gallery .
• For more information about evaluating your models, see Gen AI evaluation service overview .